CHILDS GLACIER 399 photographs by F. H. Moffit in 1906, the photographs of ice falls and advancing waves by Rex Beach and Fred Stone in 1908, and the photographs by E. A. Hegg, a professional photographer at Cordova. Beach l has written a vivid description of the glacier and river. Observations by the Railway Engineers. Between 1906 and 1910 the Copper River and Northwestern Railway was being surveyed and built, and Mr. E. C. Hawkins, the chief engineer, furnished us with a valuable body of precise data in the form of maps, photographs, and personal observations by himself and his assistants.3 This information is especially useful (a) because one of the early proposed railway routes crossed the Copper River at the southern end of the Childs Glacier cliff; (b) because the engineers remained a long time at the bridge less than half a mile east of the northern terminus, where the bridge engineer Mr. O'Neel kept precise records of river conditions and meteorological changes from 1907 to 1910, besides taking many photographs; and (c) because Mr. A. O. Johnson, one of the assistant engineers, took an especial interest in the behavior of the glaciers, and saw to it that precise maps were made during the unusually important period in the summer of 1910. To Messrs. Hawkins, O'Neel, and Johnson, and many others, we are under special obligations for turning over to us freely the results of observations between 1906 and 1910. Observations by the National Geographic Society's Expeditions. Our own observations of Childs Glacier were made in August, 1909,8 in August and early September, 1910,4 and in June, 191 l.B In 1909 we only saw the glacier in passing on the railway, but in 1910 the junior author spent about two weeks in the vicinity of Childs Glacier, making careful studies of general conditions of glaciation and of special phenomena at the time of an unusual advance. Many photographs were taken, and Mr. Lewis made the topographic map reproduced as Map 9 (in pocket). In June, 1911, four days were spent at Childs Glacier by the junior author. Normal Condition from 1884 to 1905. From 1884 to 1906 Childs Glacier is believed to have terminated in Copper River with a convex front, the amount of ice brought forward from the snowfield being about balanced at the terminus by ablation and by the discharge of icebergs that were carried away by the river. This belief is supported by such photographs, maps, and observations as are available. Allen's description of the glacier in the winter of 1885, with the river about 875 feet wide and the ice so rough that it was impossible to travel upon it, suggests a normal winter condition when lack of melting and of iceberg discharge result in advance of the ice cliff, narrowing the channel to much less than the usual summer width and breaking up the river ice in i Beach, Rei, Everybody's Magazine, Vol. XIX, 1908, pp. 778-9, 784, 785-6, 789; The Iron Trail, New York, 1913, pp. 207-208; Underwood, J. J., Alaska, New York, 191S, pp. 45-49, 68-59; Ellis, Carlyle, Overland Monthly, Vol. LVm, 1911, pp. 467-466, Technical World Magazine, Vol. XVII, 1912, pp. 3-13; Culmer, H. L. A., Deseret Evening News, Salt Lake City, Utah, Dec. 16, 1911 (illustrated by reproductions of oil paintings of Childs and Miles Glaciers). ' Before his death we had the favor of having Mr. Hawkins read and criticise the manuscript of this chapter. »Tarr, R. S. and Martin, Lawrence, Nat. Geog. Mag., Vol. XXI, 1910, pp. IS, 37. «Martin, Lawrence, Nat. Geog. Mag., Vol. XXTT, 1911, pp. 541-548; Mastering the Alaskan Glacier Barriers, Scientific American Supplement, Vol. LXXI, 1911, pp. 305-307; Gletscheruntersuchungen Ittngs der Ktlste von Alaska, Petermanns Geog. Mitteilungen, Jahrgang 1912, Augustheft, pp. 79-81, Tafeln 9, 10; Two Glaciers in Alaska, Bull. Geol. Soc. Amer., Vol. 22, 1910, p. 731. • Martin, Lawrence, in H. F. Reid's Variations of Glaciers for 1911, Jour. Geol., Vol. XXI, 1913, p. 424; Alaskan Glaciers in Relation to Life, Bull. Amer. Geog. Soc., Vol. XLV, 1913, pp. 803, 807-809.